Things we learned about social media from recording people’s use of mobile phones
Research on social media is popular. What does not seem to be so popular is finding ways to closely examine people’s actual, real-time use of social media, or in other words, how they use it moment-by-moment. So this is precisely what we did (Barry Brown and myself, with data collection conducted by Moira McGregor and Barry).
But how could you investigate this real-time use of social media as-it-happens? One approach Barry and colleagues employed is to get research participants to screen-capture their smartphone during use while at the same time recording ambient audio via the microphone.
The result is a richly detailed set of data captured from everyday situations where you can see considerable detail of how people interact with their smartphones, browsing Facebook, Twitter or Instagram (to name just three social media systems). Since smartphones are pretty much always with us, you can also hear how participants chat with others as they use their phone. It enabled us to develop a detailed understanding of this interaction, supported by transcription of talk and the visibility of on-screen action.
Here’s a figure from our paper to illustrate:
I’ll return to this example in more detail next. Now, here’s three things we learned from looking at this data.
1. Social media use is expansive.
In the data we often find the use of social media being brought to bear as an interactional resource during the everyday moments. What this means is that social media use itself can move face to face conversations on, such as through jokes or introducing new topics to talk about. What does this look like? Here’s a simple example where A and B are chatting about a Facebook post that B has just made. B posts “Whee!” as a status update which A then sees and then transforms into a question for B (“we?”, as illustrated in animated GIFs below!):
B is a bit confused by this:
B is still confused by A’s question, but A persists (“you said WE”); B essentially verbalises his status update with a high-pitched “oh, wheeee!”, producing a groan of acknowledgement from A (“oh my gawd!”):
So what? In essence this suggests an alternative account to the popularly-held view that the use of social media, and smartphones use in general, constitutes a unique distraction from normal ‘morally desirable’ forms of social interaction (a view espoused by Sherry Turkle). Such a view is simply not borne out in our data.
The above example also illustrates the next point well.
2. Interaction with social media is finely and intricately interwoven with the proceedings of everyday life.
In the previous example, what A posts on Facebook occasions a brief humorous verbal interaction with B. If you look carefully, how B scrolls around his news feed fits in with their conversation. In short, it’s interwoven.
There’s a clearer example we can draw on. Here we join three people (a different A, B and C) while C is browsing Facebook on his smartphone. C is taking part in a conversation with A and B. What we draw attention to is just how finely and sensitively coordinated C’s talk is to this conversation and his use of Facebook.
C is scrolling through photos on Facebook. A and B are having a conversation about science fiction and fantasy books. Just as B finishes talking (“don’t they just have like sci-fi and fantasy books”), C takes advantage of the fact that there is a pause in the photos loading to interject. He breathes in and talks (“there no there’s…”) just as A also starts talking (A: “but- that’s the thing is that…”):
The point here is that C is not just ‘browsing Facebook’ but doing so in a way that is synchronised with the ongoing conversation he seems to be part of. He takes advantage of the slowness of Facebook to jump into the conversation at an opportune moment.
That interweaving takes place in some sense is ‘obvious’, but it actually has really significant implications. It suggests that we must rethink where meaning of social media is ‘located’ when we seek to study its use. Meaning is not necessarily to be found ‘in’ social media at all, but rather in the interactional process of ‘gearing in’ the use of social media with the mundane, deeply practical contingencies of everyday life.
In this way our research suggests that the picture gathered from either large-scale aggregations of social media data or post-hoc interviews of users may be missing something. But more about that later.
3. There’s a great similarity in the methods people use to interact on social media and those we use in everyday verbal talk.
By ‘methods’ we mean the familiar ones we all employ on a daily basis, e.g.: taking turns to speak, repairing one another’s utterances, selecting who to speak next, and employing common patterns of adjacent pairs of things (e.g., question/answer, summons/response, greetings, etc.). These normal methods are ‘tweaked’ to fit the design features of social media. So, for instance, social media users employ the ‘@’ feature to address others, but in ways that are /a bit different/ to how we preface an utterance with someone’s name when we select them as the ‘next speaker’. For instance, a lot of ‘@’ use also is used to sort out who is ‘in play’ and whether they are considered to be a relevant ‘next speaker’ at all.
The wider significance of this is that it means a whole gamut of concepts and findings from ethnomethodology and conversation analysis (or ‘EMCA’–which has extensively studied naturally occurring talk) can usefully be applied to help make sense of the complex interactions people perform on social media.
Question 1: But isn’t there lots of social media research already?
But where does this all fit into existing social media research, of which there is plenty?
When we looked at what research was out there on social media, we found that most of it seems to fit into a couple of two overlapping categories. We called these ‘actor-focussed’ and ‘aggregate’ perspectives in our paper:
- ‘Actor-focussed’ perspectives are mainly about eliciting users’ post-hoc accounts of their behaviour on social media. So, interview studies, surveys (e.g., questionnaires) and so on.
- ‘Aggregate’ perspectives are probably more popular, and involve scraping data from social media itself. So, Twitter postings, friend network mapping, etc.
Yet productive as these approaches have been, what ends up being absent from them is an examination of how social media concretely features in the mundane ‘everyday world’ of its users; this is the kind of approach we have taken in the points made above.
Our inspiration for this different approach is ethnomethodology and conversation analysis. You can read more about this particular perspective in the paper. Suffice to say here that it relentlessly prioritises close inspection of the details of how human action (e.g, with technology) is organised.
Question 2: Does this kind of research into social media tell you what to design?
The simple answer is no, not directly.
But what we do find is that looking at moment-by-moment use throws into relief some of the design choices that have been made. Take commenting on social media for example. Most systems hide what it is you are typing until you hit ‘post’. This means that there are many ‘conversational’ things that cannot be done, such as someone repairing what you are saying as you say it, or enabling someone else to quickly respond to what you are saying as you say it. In other words, online chat becomes less collaborative and slower.
You can read our CSCW paper here.
Originally published at notesonresearch.tumblr.com.
